IIT Madras Wadhwani School of Data Science and Artificial Intelligence’s (WSAI) Centre for Responsible AI (CeRAI) is focused on advancing Al governance in India and the Global South while ensuring it remains safe, trustworthy, fair, and inclusive.

Towards this, the Centre organized the ‘Conclave on AI Governance’ at the IIT Madras campus today (7th Oct 2025).

This Conclave on AI Governance is an official Pre-Summit Event of the India-AI Impact Summit 2026, an ambitious, future-focused global gathering announced by Shri Narendra Modi, Hon’ble Prime Minister of India, at the France AI Action Summit in February 2025.

Shri Abhishek Singh, CEO, IndiaAI Mission, who was featured on the TIME100 AI 2025 list for significant contributions to democratizing AI services, especially in rural India, delivered the keynote address

The Conclave also witnessed the launch of path-breaking new initiatives and pioneering studies including,

1. A report titled “The Algorithmic–Human Manager: AI, Apps, and Workers in the Indian Gig Economy”
2. A discussion paper titled “AI Incident Reporting Framework for India”
3. Policy chatbot called “PolicyBot,” an open-source, interactive system designed to make complex policy and legal documents accessible to non-experts through reliable question-answering
4. IndiCASA Dataset - Centre for Responsible AI (CeRAI), WSAI, IIT Madras is launching a dataset that can be useful for bias risk detection and assessment in LMs in the Indian context.
5. AI Evaluation Tool - The objective of the evaluation tool is to provide a unified and automated framework that evaluates Conversational AI systems in a consistent, transparent, and scalable way
6. COIN Network: The Co-Intelligence (COIN) Network, established in collaboration between the Wadhwani School of Data Science and AI (WSAI), IIT Madras and itihaasa Research and Digital, is a global network of individuals and organizations dedicated to the pursuit of leveraging “co-intelligence” to benefit enterprises and society

The event saw not only academics, researchers and students from international and Indian institutions but also policymakers and industry leaders deliberating on three key thematic areas -

⮚ Safe and Trusted AI: The discussions centered on AI self-regulation through voluntary commitments and AI incident reporting. This session focused on enabling the developers and deployers to efficiently manage and mitigate risks and harms.

⮚ Inclusion: This session explored opportunities for adopting participatory AI frameworks and how ensuring fairness could lead to more inclusive AI systems that addressed the needs of diverse socio-economic contexts and unique user perspectives.

⮚ Human Augmentation: The focus was on the “Co-intelligence revolution,” a human-centered approach where AI augments rather than replaces human skill sets and agency.

Addressing the conclave virtually, Shri Abhishek Singh, Director General, National Informatics Centre (NIC), and Additional Secretary, Ministry of Electronics and Information Technology (MeitY), Government of India, said, "It requires the country, the global South, to get a voice, to get a place at the high table of AI in order to ensure that we are able to come up with a framework which is truly inclusive, which is truly participative. We have consistently championed a multi-stakeholder model for governments, involving not only governments but also industry, academia, and civil society. The Global Partnership on Artificial Intelligence is one such party stakeholder body, the UN High Level Advising body on AI was also a very representative body on which we could come up with common frameworks around that.”

Further, Shri Abhishek Singh, also added, "What we also hope is that the framework that India is defining for our AI governance can become a role model for countries of South and we are also a part of the summit. We are engaged in discussions about how we can support similar activities in these countries, in what ways we can augment the efforts that are going on in these countries, in what ways we can create a deposit of AI, AI tools which can be developed and deployed in these countries and in addition. As part of the AI impact summit, we are doing several pre-events for the impact summit, and the idea is to again spread awareness and this message with regard to using AI in a responsible manner. So I'm sure the event that IIT Madras is hosting today, the Conclave on AI Governance will contribute a lot towards this discussion and to this discourse."

Shri Abhishek Singh has over 25 years of experience across public administration and policy, technology for governance, and driving digital transformation initiatives at scale.

Delivering the inaugural address, Prof. V. Kamakoti, Director, IIT Madras, “As we start growing any technology, we have two sides - good and the bad. There is the thin line that we need to draw to bring out certain regulations that will not kill the technology but prevent its misuse. That is what we are attempting to do. One thing that is going to be constant in this process is change and how to accommodate this change is going to be the biggest challenge in the Governance framework.”

Further, Prof. V. Kamakoti added, "When it comes to certain decisions that are critical from a life and human point of view, in fields such as healthcare and justice systems, where AI is going to be used, these points are going to be deliberated. Bias is not just a mathematical problem. How do we define good? This is going to be an inter-disciplinary effort. I see how humanities is going to have a major impact."

Prof. V. Kamakoti said, "I would like to teach Indian pedagogy and Indian Point of View to our children. If we are going to recommend a chatbot for education, how do we certify it? Should we have a 'Censor board' to state what should be fed in and what should not be fed in. This is going to be an interesting question and an important factor."

Highlighting the need for such conclave to guide AI Development responsibly in India and abroad, Prof. B. Ravindran, Head, WSAI, IIT Madras, “AI technologies present a varied landscape and hence require a multi-pronged approach for effective governance. This has to involve voluntary commitments from service providers as well as active participation of various stakeholders and end-users to create a valuable Human-AI ecosystem. This conclave addressed some of these aspects and we hope will form a valuable input to the discussions leading to the AI Impact Summit.”

KEYNOTE SPEAKERS

The conclave witnessed multiple high profile keynote speakers including Prof. Srinivasan Parthasarathy, University Distinguished Scholar, Computer Science and Engineering Department, Ohio State University (OSU). He directs the data mining research laboratory at OSU and co-directs the university-wide undergraduate program in Data analytics. He is a recipient of an Ameritech Faculty fellowship in 2001, an NSF CAREER award in 2003, a DOE Early Career Award in 2004, and multiple grants or fellowships from IBM, Google and Microsoft.

Bringing a global perspective on the future of AI-driven research, Prof. Srinivasan Parthasarathy said, “The integration of AI and advanced analytics into society necessitates the responsible management of the multifaceted risks involved. This challenge is socio-technical, requiring thoughtful, participatory human involvement at every stage. Lessons from critical, high-uncertainty, and complex domains reveal that joint human–AI systems consistently outperform either humans or machines alone. When designed effectively, such collaborative human-machine teams often leverage their complementary strengths, resulting in safer, more resilient, inclusive, auditable, adaptive, and trustworthy outcomes.”

Other keynote speakers were Prof. Venkatram Ramaswamy, Professor, University of Michigan and Ms. Urvashi Aneja, Founder and Director of Digital Futures Lab, an independent, interdisciplinary research studio that studies the interplay between technology and society in the Majority World.

Ms. Urvashi Aneja is currently a member of the UN Committee on Safe DPI, the Government of Telangana’s Committee on Responsible and Ethical AI, the Expert Advisory Committee for the Global Index on Responsible AI, and the advisory board for Quad 9. She is also an associate fellow at the Center for Responsible AI at IIT Madras. She featured in an MIT Sloan Management Review India's Policy 50—a definitive list of 50 leaders steering and shaping the nation’s digital economy.

Prof. Venkatram Ramaswamy is the Hallman Fellow of Electronic Business and Professor of Marketing at the Ross School of Business, University of Michigan. A globally recognized thought leader, his research blends innovation, strategy, branding, operations, and human-centric value creation. He co-authored The Future of Competition (with C. K. Prahalad), introduced the concept of co-creation, and has published landmark work on experience innovation and building co-creative enterprises. He is also a distinguished faculty fellow at WSAI.

Panel Discussions were also held on the following topics:

⮚ Implementing voluntary Frameworks for RAI in India

⮚ Participatory AI - Need for end user/community engagement to build fair and inclusive AI Systems

⮚ Thriving in a co-intelligence World - NEST Ecosystem

BRIEF ABOUT THE REPORTS / INITIATIVES LAUNCHed

1. A report titled “The Algorithmic–Human Manager: AI, Apps, and Workers in the Indian Gig Economy

A report titled “The Algorithmic–Human Manager: AI, Apps, and Workers in the Indian Gig Economy”, by the Centre for Responsible AI (CeRAI) at IIT Madras, examines the profound impact of Artificial Intelligence (AI) and digital technologies on the burgeoning blue-collar gig economy in India. Through a mixed-methods approach, including in-depth interviews with gig workers and key stakeholders, the report uncovers a dual reality:

A. While AI and technology create unprecedented access to work and new efficiencies, it also introduces significant challenges related to fairness, transparency, and worker dignity.
B. While AI is responsible for the algorithmic management of gig work (allocation, performance management, etc), workers believe that a human is making these decisions.

The report advocates for a "pragmatic middle path" that fosters innovation while establishing robust digitally-enabled guardrails to protect workers

2. A discussion paper titled “AI Incident Reporting Framework for India”

The Centre for Responsible AI (CeRAI) at IIT Madras has published a discussion paper proposing a tailored framework for AI incident reporting in India. The paper presents a detailed roadmap for defining, identifying, and managing AI incidents in the Indian context. It recommends creating a nationwide federated database of AI incidents and a centralised, democratised AI incident reporting portal.

The framework recommends a federated database and a centralised reporting portal, outlining a six-level incident management process and a governance structure. It also advocates for a multi-dimensional AI risk and harm taxonomy and context-aware mitigation strategies specifically for AI systems in India

3. Policy chatbot: “PolicyBot” - Policy and legal documents are central to governance, but remain difficult for non-experts to navigate due to their length, complexity, and specialized terminology. PolicyBot is an interactive system designed to address this challenge by enabling reliable question answering directly over policy documents.

Unlike generic QA systems, PolicyBot ensures that every answer is grounded in the source text, allowing users to expand and verify the exact passages that support each response. The system is fully open-source, containerized for reproducibility, and deployable on modest hardware using a local-first strategy, thereby preserving privacy and enabling its use in bandwidth-constrained environments.

4. IndiCASA Dataset - Centre for Responsible AI (CeRAI), WSAI, IIT Madras is launching a dataset that can be useful for bias risk detection and assessment in LMs in the Indian context. The new dataset IndiCASA (IndiBias-based Contextually Aligned Stereotypes and Anti-stereotypes) dataset addresses a critical gap in artificial intelligence: the inability of global models to capture the nuanced societal biases specific to the Indian sub-continent.

IndiCASA offers around 2.5K human-validated sentences designed to include stereotypes and their anti-stereotypical counterparts across five major demographic axes: caste, gender, religion, disability, and socioeconomic status. This extensive, culturally rich collection was curated using a "Human-AI Collaborative" methodology, leveraging initial sentences from the existing IndiBias dataset, generating variations using advanced LLMs, and finally submitting all content to review by Linguistics and Social Science experts specializing in Indian demographics

5. AI Evaluation Tool - The objective of the evaluation tool is to provide a unified and automated framework that evaluates Conversational AI systems in a consistent, transparent, and scalable way

The objective of the evaluation tool is to provide a unified and automated framework that evaluates Conversational AI systems in a consistent, transparent, and scalable way. With it, the researchers compare the AI systems fairly, measure performance across multiple facets, and build trust in AI deployment. The tool connects directly with conversational AI agents to simulate human interaction. The AI agents are subjected to structured test plans, apply evaluation metrics and strategies, and finally generate a comprehensive evaluation report. This ensures the process is systematic and repeatable

The tool simulates human interaction to subject AI agents to structured test plans, measuring performance across key facets like accuracy, robustness, and fairness. Its Version 0.2.0 includes seven test plans, 48 major metrics, and over 400 test cases across several Indian languages to ensure consistent and transparent evaluation.

6. COIN Network: The Co-Intelligence (COIN) Network, established in collaboration between the Wadhwani School of Data Science and AI (WSAI), IIT Madras and itihaasa Research and Digital, is a global network of individuals and organizations dedicated to the pursuit of leveraging “co-intelligence” to benefit enterprises and society

It explores five key futures—Planet (Nature), Development (Society), Work (Economy), AI (Technology), and Learning & Governance.

The initiative is being led by Prof. B. Ravindran, Head of WSAI, IIT Madras and management guru, Krishnan Narayanan, Co-founder of itihaasa Research and Digital and includes global thinkers such as reinforcement learning expert such as Prof. Venkat Ramaswamy, Ross School of Business, University of Michigan, Indian tech-historian, Vishwanathan Anand, Chess Grandmaster and World Champion, and others


ABOUT WADHWANI SCHOOL OF DATA SCIENCE & AI:

IIT Madras has made several impactful research contributions in Data Science and Artificial Intelligence over the past year, with its work featured at premier conferences worldwide. These achievements are powered by the Wadhwani School of Data Science & AI, which is at the forefront of several leading research centres and labs: the Robert Bosch Centre for Data Science and Artificial Intelligence (RBCDSAI), the Centre for Responsible AI (CeRAI), the Centre for Integrative Biology and Systems mEdicine (IBSE), AI4Bharat, the Walmart Centre of Tech Excellence, the MInT Collaborative, and the Alpha Grep Quantitative Research Lab. Researchers from these groups have produced groundbreaking work and have made significant contributions to the DSAI ecosystem.

FOLLOW US ON LINKEDIN / X